Diagram-MMU - Code Editing: leaderboard
Metric: All Score (%). Source: vi-ocean.github.io. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 VL 235B A22B | 52.25 |
| 2 | Kimi K2.5 | 52.13 |
| 3 | Gemini 3.1 Pro (Preview) | 51.22 |
| 4 | Qwen 3.5 397B A17B | 50.82 |
| 5 | GPT-5.2 | 46.15 |
| 6 | Seed 2.0 Pro | 43.73 |
| 7 | InternVL3-38B | 38.85 |
Interactive version: theaggregate.ai/benchmark?slug=diagram-mmu-code-editing · How It Works · Data refreshed daily, snapshot 2026-09-05.